Partial Response Equalization With DFFE Pre-Cursor ISI Cancelation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing decision feedback equalization techniques for partial response equalized signals face challenges in handling pre-cursor interference and require complex parallel architectures, leading to high computational complexity and inefficiencies.
Innovation Solution
Applying decision feed forward equalization (DFFE) with pre-cursor cancelation capabilities to partial response systems, utilizing a partial response inverter to invert symbol estimates and iteratively cancel both pre-cursor and post-cursor interference, while maintaining causality and reducing complexity through reuse of estimates across stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If decision feedback equalization (DFE) is used to mitigate intersymbol interference, then ISI cancellation capability is improved, but device complexity grows linearly with channel memory and requires parallel architectures that grow exponentially with the number of unrolled taps
Solution Approach 1:
The patent segments the equalization process into two distinct parts: a feedforward filter that processes current and future symbols, and a feedback filter that processes past symbols. This segmentation allows independent optimization of each filter's complexity, reducing the overall computational burden compared to traditional DFE which processes all symbols through a single complex feedback loop.
Solution Approach 2:
The feedforward filter performs preliminary equalization by canceling pre-cursor ISI before the feedback filter processes post-cursor ISI. This preliminary action reduces the complexity of the subsequent feedback operation, as the feedback filter only needs to handle residual interference rather than the full ISI burden.
2Speed
If parallel architectures such as loop unrolling are used to overcome the bottleneck created by the decision feedback loop, then processing speed is improved, but device complexity grows exponentially with the number of unrolled taps
Solution Approach 1:
By segmenting the equalization into feedforward and feedback components, the patent avoids the need for extensive loop unrolling. The feedforward filter can be implemented with simple parallel operations for high speed, while the feedback filter processes sequences more efficiently, together achieving high processing speed without exponential complexity growth.
Solution Approach 2:
The patent employs adaptive filtering where the feedforward and feedback coefficients are dynamically adjusted based on channel conditions. This dynamic adaptation allows the system to maintain optimal performance across varying channel states without requiring fixed complex parallel architectures for all possible scenarios.
3Reliability
If traditional DFE is applied to partial response systems, then post-cursor interference cancellation is improved, but pre-cursor interference remains unhandled and requires additional complexity
Solution Approach 1:
The patent explicitly segments interference cancellation into two functions: the feedforward filter handles pre-cursor ISI cancellation while the feedback filter handles post-cursor ISI cancellation. This segmentation eliminates the need for additional complex processing stages, as each filter type naturally addresses its designated interference direction.
Solution Approach 2:
The feedforward filter serves multiple purposes: it equalizes the signal and simultaneously cancels pre-cursor ISI. The feedback filter similarly serves dual functions of signal refinement and post-cursor ISI cancellation. This multi-functionality reduces overall system complexity compared to separate dedicated circuits for each function.
Data Source
AI summary
A receiver includes a partial response (PR) system that receives a received signal from a transmitter over a channel and equalize the received signal such that there is a controlled relationship between consecutive values of equalized received symbols and transmitted data transmitted by the transmitter. The receiver also includes a decision feed forward equalization (DFFE) system that receives partial response signals from the PR system and cancel at least one of pre-cursor intersymbol interference (ISI) or post-cursor ISI introduced by the channel.